Jump to content

Turbocharger

Settled In
  • Posts

    2,781
  • Joined

  • Last visited

  • Days Won

    2

Everything posted by Turbocharger

  1. I put a Mitsubishi VNT onto a 300Tdi more than ten years ago, and spent some time tuning it. You can eliminate the spiking in normal use by making it more sluggish to respond. I've not seen an issue from brief overboosting (since it won't be adding enough fuel to make power it's just pressurising the intercooler and arguably the cylinders, but with excess air so cooler). Aside from the gauge rattling the end stop, what's the problem you're trying to solve here?
  2. Quick one - does anyone know the spec (thread and length) of the two bolts which attach the handbrake adjuster on the back of the LT230 (Defender/Disco/RRC)?
  3. Same as Fridge, I've got a stack of temperature stickers scattered around sumps, header tanks etc. There's little I'd do in real time to react to them, gauges are just cab ornaments really.
  4. This is exactly what I did - I have a Mitsubishi Industries VNT on my 300Tdi, there's a build thread somewhere. Edit: A friend made up a manifold, I moved the pressure diaphragm to the opposite side and played around with a custom boost capsule and swapped in different springs, preloads etc to get a nice proportional operation. I put quite a bit of science into it, I think the pics have disappeared but the car is here so I can look up any details. In the end I think I used a standard 300Tdi wastegate actuator - it's effectively preloaded to 1.2bar and then opens fully by 1.4bar; that puts the turbo in the loop and it'll settle under constant boost in any given position according to the massflow. So long as the vanes aren't adjusted too "closed" on the endstop it'll go fine, just listen for early surge when you're being aggressive with settings and throttle.
  5. I think it's fair to say I'm a year in to this project and progress is slow, it's been a cold winter and the damn thing won't build itself. If the OP wants to lower the car for big tyres but still fit in the garage, that's fine, we're a broad church. Biggest you can buy for those rims is probably 255/85R16 or 285/75R16, you can search blackcircles.co.uk etc by tyre size to see what's out there. 18s are available for Defenders which might give a different choice, especially in a road pattern. You may need to extend the bump stops and reduce the steering lock to stop larger tyres from rubbing though.
  6. I bought a suspiciously cheap WiFi camera off eBay, and it works "some of" the time, but I wouldn't buy again. Raspberry Pi isn't an easy answer for live video either, it seems. My camera is a £30 WiFi eBay job, pair your phone to its network and then the Chingrish app shows a picture, unless it's in a bad mood and then it shows 'connecting' regardless of the number of restarts. Even when you're out and about, connecting to its WiFi obviously doesn't have a data connection, so no email or surfing etc. At home it won't preferentially swap to the camera network when it sees it, so no use for reversing onto the driveway. Most unforgivably, there's sometimes a lag in the video stream, so what *was* a gap to the trailer/wall/other car is suddenly a dull thud. I think Pi is the answer here if you just want to use your phone, but the data connection/WiFi issue isn't trivial.
  7. Loving Meccano's approach. I've had a fallow few months because it's really cold outside(!) so I've been learning more about Fusion360 and getting better at 3D-printing stuff. It can dangerously become the answer to everything though, but I've found an economic source of MIG gas again too so we may be able to add metallic fabrication back to the arsenal soon. My progress has taken much longer than the photos suggest - essentially I've put the electronics in a box and made three brackets; I'm hoping this is the bit where lots of effort doesn't bear much fruit. 3D printing brings a nice custom finish to this, though the jury's still out on whether it'd be adds value or if it'd be quicker to just buy something or fold up some metal. Hoping-it's-temporary box for the sparky bits: And a few clamps to hang the solenoids off the cage to stop it wandering around. Next to do is extend the solenoid loom and height sensor looms to reach the dashboard and put a neat little plug on them. I don't have the patience or steady hand for soldering so I've ordered a third-hand clamp thingy.
  8. These are PLA. I've got a reel of ABS once I've finalised the design, but if it's working I'll probably just replace on failure. A friend runs a printer farm for a living so he's very aware of the failure rate. One of his machines is 6ft tall and has print times longer than a week, so regular checks prevent the spaghetti mess. Interested in your adapter - I'm just using a vibration isolation ring and some Velcro to hold the top in place for the Disco2 springs - I've a bayonet at the bottom and the travel will be restricted by the shock so I'm not sure why it needs pinning mechanically at both ends.
  9. Finally back on this - Corona brought other priorities. I've been loaned a 3D printer by a friend and (after learning how to use it and tune the prints) I can make the designs I want, recognising the amazing amount of time that repeating metalwork fabrication takes up - especially if I want a decent result. I've printed the height sensor setup for the nearside rear. I really don't want to drill the chassis as it's galv, so better than being a "bolt-on" conversion, I can do custom firtree plugs for existing chassis holes so it's just a push fit. Using a clip arrangement here for the height sensor so I can use the P38 parts, even if the mounts have snapped off. Of course, now it's in CAD I can just flip the design 180° and print the other side, not back to doing battle with the miller. Here's a video in action: There's much more fabrication to get the air tank and solenoids mounted, but this feels like progress at least.
  10. Does this mean that the "Mercedes" manifold could (be made to) fit the 300Tdi head?
  11. A little more time spent building a programme structure (and chasing issues with code), I'll use the electric window switches either side of the 2002 dashboard as control inputs in the cab. They'll be nominally MENU/SELECT on one side and UP/DOWN on the other. I've written a really simple routine which inflates or deflates each solenoid, and shows ^ or v on the display in the relevant corner. I can also equalise an axle left to right or one side front to back (open the axle solenoids, don't inflate or deflate), and here it puts = in the relevant corner of the display. It's useful to be able to manually ping each solenoid but normally it'll set a target height for each wheel and have a routine which does the legwork. As it can only inflate or deflate at one time the routine will look at all four corners, decide whether the average is higher or lower than target, then run the relevant routine to inflate or deflate the appropriate wheels and show the change on the display. That's all for the next stage though, I haven't mounted height sensors yet. Right now I've got a controller where MENU takes you through seven modes, and SELECT will do different things in each mode to choose what UP/DOWN do. Automatic - will set a target height and adjust to meet that, plus or minus a tolerance band. Up/Down will adjust the target height, then the control routine does its thing to best match that, and then rechecks. Lean - takes the target height from Auto, and up/down will offset the targets left/right. Pitch - same, offsets front to back instead. Calibrate - this will be to manually set the min and max heights for corners, rebalance one dicky corner sensor etc. Test - this tests the solenoid outputs, in a pattern, individually, pairs together etc, controlled by UP/DOWN. This should help with any diagnostics on the trail, listening for the click etc. Diagnostics - will page through different inputs and variables on the display using SELECT, then UP/DOWN. Manual - this uses SELECT to page through different options (all, front, rear, left, right, FR, FL, BR, BL, equalise across, equalise one side) and then uses UP/DOWN to activate appropriately. At the moment only modes 5 and 7 are written - the others need height sensors (or me to have some patience...). This does mean I can control the solenoids in a meaningful way to make it dance around a little: And an explanation of the controller, mostly for myself in lieu of formal documentation: Some pictures in action from manual mode (you get the idea): Inflating the rear: Equalising the right-hand side springs: Deflating just the front right bag: Next steps are to tweak the spring seats which I'm not happy with. to check that the shocks don't allow the springs to overstroke (even if the controller starts exhausting them as the axles droop). to fit 4x height sensors so that the control can move to automatic.
  12. Update: I've spent an embarrassing number of hours with a soldering iron making a hash of some Veroboard and I have a working interface for the solenoid driver. I'm using an Arduino Mega Pro (54 digital IOs and 16 analogue IOs - for £12!), and the electric spaghetti is to make the 5v signals into 12v signals. After a smoky discovery that the heat sink tabs are electrically connected to the transistor outputs AND earth (oops) I need to rethink that area (see the paper scraps in the video), and sort out a couple of dead channels, but the thinking is fundamentally sound - it clicks! (Just on/off at this stage to make sure it works). There are 8 channels on the driver - four wheels, inflate, deflate, compressor and a spare so I can swap wires around if there's a problem on the trail. I'll also have manual push buttons to force 12v directly, and 4x inflation points as another level of backup. Next steps are to sort out the input wiring (dash switches, inclinometer, GPS, Bluetooth etc) and a dash display - if I get all the hardware soldered in place I can work up the code gradually from 'minimum viable product' to something clever, rather than delaying perfection and becoming the archetypal vapour project. Then I need spring mounts, shocks, height sensors and looms - and an MOT!
  13. Waiting for some electronic bits to turn up now, so I whizzed up a couple of lower spring plates on the mill. These are temporary (probably) but will support the lower spring seat and could be faced at an angle to help offset the effect as the axle rotates on full droop, which is tending to 'peel' the lower spring seat out of the bag. The mill is new to me but pretty pleased with how this turned out, the bolt pockets neatly the right width to make the bolt head captive too. (The speed holes around the edge are definitely to reduce unsprung weight and to look cool, nothing at all to do with using old lumps of alu from the offcuts bin at all, oh no). I've decided to built the control as a daughter board onto an Arduino Pro Mega - essentially an Arduino chip, but the Mega has 16Mhz processor, 54 digital IOs and 16 analogue channels - that should be plenty even if I decide to put accelerometers on each wheel! The Pro Mega puts this onto a PCB which is 55 x 38mm and has pins at 0.1" spacing for use with breadboards - and all for £7, which is the really impressive part!
  14. More electric spaghetti but it looks more complicated than it is - the Arduino voltage regulator and the solenoid driver are now on the Veroboard and I'm using a real height sensor. The circuit just needs duplicating for the other solenoids and probably another for the compressor. (I could even drive the compressor at variable speed depending on the pressure in the tank, but that's probably one for v5.0!) Much more boring work to do, little tasks that don't look like big progress - then suddenly with one more job it'll look like progress again.
  15. One of those productive days with something to actually show for it. I've got a bench version of the air system, an air tank and a battery to keep it all alive. The compressor will feed the system (slowly) and has an automatic unloader valve, but I think more wires would be required to make it cut off at max pressure. With care, that's not an issue here, and I've got a T piece to top it up with the garage compressor. I've got Inflate, Deflate and four wheel solenoids. The RR valve block and driver makes it very easy, just 12v to the right places and the car goes up and down. A good chance to play with the springs on the car - curtailed when something was binding, then released and popped the spring, which lobbed the (unbolted) lower spring mount somewhere near the opposite corner of the car. Still lots of little fabrication jobs to do (plus coding and height sensors, wiring, piping etc etc), but there's much less to show for those hours in the garage!
  16. That one is just a piece of box with three overlapping holes cut into it, but I'm not happy so v2.0 is on the milling machine for a tidier result. I think that's my question - it's no 'safer' if they're bolted to the chassis or not when it comes too far apart, so therefore it doesn't matter? A check strap (or shorter shock) is a backup option, my shocks are fairly long to maximise the travel. Shocks are slightly outboard of the bags so straight 'heave' (rather than articulation) should be the worst case.
  17. A month on, and a little progress between other distractions. I have a little MOSFET circuit to take 5v from the Arduino to 12v to drive the solenoids at the high-side. It's specced to drive the compressor too, if it's needed further down the line, but for now the unassuming 'click' with the red LED shows that I can drive a big voltage with a small voltage - I'm learning something, at least. The potentiometer mimics a height sensor, the green LED is for 'inflate' and red is for 'deflate', which is the solenoid that's clicking. https://youtu.be/0a5vEektG5k I've also got a fixing for the rear airbag - fronts are on the milling machine now, and I'll remake the rears to tidy them up I think. https://youtu.be/cbXoZAI3StA I'm keen to secure at least one end of each bag but they'll fail if the two ends come too far apart so they need to be restrained by the shock travel, regardless of whether they're pulled apart mechanically or pushed apart by the internal pressure, so I could mirror the security of the coils and only hold one end. Any thoughts?
  18. Thanks - nothing more complex than holes in the chassis and a tab on the trailing arm then. I can do that 👍
  19. Perfect, thank you, that's what I'd expected. Don't suppose you've got any rear height sensor pics too? I can be creative, there's more space back there, just handy to see how the factory boys did it.
  20. Thanks Paul - you have a PM. Wes - I hadn't appreciated it would be so easy - from the hit&drop info I'd planned to code up the PWM bit, but it does look like I can just power up the solenoid driver on pins 10/11/12/13, then put 12v to pins 1-6 for whichever solenoids I want to play with. Thank you - that'll make the coding easier! Still needs MOSFETs to interface logic to 12v but less need for the inductive flyback protection etc. Does anyone have an in-situ photo of how the height sensors were mounted on a RR Classic? I think that's what the little 'ear' on top of the radius arm is for just behind the axle (but I'll bet it's a person I'm not that keen on to drill!)
  21. For general vehicle design info (generally a 'how it's done' for modern and electronic systems rather than a first principles dynamics approach) I'd recommend the Bosch Automotive Handbook: https://www.amazon.co.uk/Automotive-Handbook-Robert-Bosch-GmbH/dp/1119530814/ref=sr_1_1?dchild=1&keywords=bosch+automotive&qid=1592901267&sr=8-1 10th edition is the latest, it's a bit spendy - you might find an earlier edition at a better price if you don't need the latest on aftertreatment OBD control etc.
  22. That'd be handy. I've seen the hit&drop graph on the P38 sites but some pinout info would be gold. I'll still use some MOSFETs to drive it but it makes the coding easier. Yes, ish. The closed length is still too great but it's got the lower coil seat upside down as a temporary carrier, so there's about 40mm to be had there. I can happily spin off the mounting lug inside the lower seat in a lathe, and/or shape the spring chair on the axle if necessary. I'm aiming for minimum mods though, and not just because I'm lazy...
  23. Only a little progress recently, kids + nice weather, and some family things over the last few weeks. Thanks for the plug info, much appreciated 👍 Any idea on these plugs, the electric window switches in a 2002 Defender dash (and a Disco 1 centre console, and late RR Classic - and let's be honest, probably a Metro or Marina too). They look like an elegant way to get inputs that will be glove-appropriate. Today I've managed an hour for some sundry bits, including the bashing and filing to make the Paddock dual spring mounts fit a standard car. Good news: it looks like it'll package nicely. If anybody knows what inputs and outputs it uses, I'm all ears! No point reinventing the wheel/transistor, though starting from scratch gives me full control of what's going on.
  24. Took a couple of weeks away with various lockdown complications... I decided I might do more if it's indoors (and that's half the point of the project anyway). It does *technically* fit indoors on the bump stops, but I snagged a sharp edge on the highest point of the tilt 🙄 Nothing more on the air yet...
×
×
  • Create New...

Important Information

We use cookies to ensure you get the best experience. By using our website you agree to our Cookie Policy